Conventional Web-based form-rendering systems, such as a server device that renders an HTML form for requesting client devices, generate a logical representation of a form as a collection of objects each time that the form is requested. A form can expose an object model that allows a form designer to write code that processes events generated while a user edits the form via a user interface displayed at a client device. Additional objects may also be generated for each form request from multiple client devices. For example, some objects are specific to each editing session which is the time duration from when a user at a client device begins to fill-out a form until the form is closed.
A server system, which may include any number of form servers, may process hundreds of concurrent requests for a particular form. Server system resources incur undue overhead when having to recreate the objects every time that a client device submits a form request. Additional overhead is required to maintain all of the private objects for the duration of each session.
Systems that implement an architecture where objects are created and destroyed for every request are inefficient, and overly-tax server resources, such as processors and memory. In addition, these types of architectures do not scale in a desirable manner and may require complex synchronization.